56
J. R. KELLER, E. MATIJEVI~ AND M. KERKER
Vol. 65
HETEROPOLY COMPOUNDS. VI.’ FURTHER STUDIES ON BASICITIES OF SOME HETEROPOLY ACIDS2 BY J. R. KELLER,* E. MATIJEVIC? AND M. KERKER Clarkson College of Technology, Potsdam, N. Y . Received M a y 9. 1060
Using a spectrophotometric technique described previously4 and the indicators, brom cresol green, methyl red and brom cresol purple respectively, the basicities of these several heteropoly acids were determined in highly diluted aqueous solutions: Bmo&bdochromic( 111), l%tungstosilicic, 9-tungstophosphoric, 12-tungstophos horic and lZmolybdoceric( IV) acids. The basicities are in agreement with these several formulations of these acids : H&Mo~OZI, H~S~WISOU, H~PzWI~O~Z, HrPWIr04tand H & ~ M O ~ ~ The O ~ latter ~ . two acids become fully ionized only at very low concentrations (6and 4 X lo-’ M, respectively). It waa also shown that the 1Ztungstophosphate ion absorbs differently from the undissociated acid in the ultraviolet, so that it actually behaves as an acid-base indicator.
Introduction
phosphoric acid, 12-tungstosilicic acid and 6In an earlier paper4we have presented data on the molybdochromic acid with these indicators in order basicities of several heteropoly acids using a spec- to test the stability of their ions to hydrolysis of trophotometric method based on the ionization of the type discussed and also to provide a check methyl orange under the influence of added acids. on technique. In addition, we have found that the optical denThe concentration range for which the acids may be investigated depends on the p K of the indicator sity of 12-tungstophosphoric acid solutions is employed. Using methyl orange we could estab- strongly p H dependent in the near ultraviolet. lish that the 6-molybdocobaltic(III) and 6-molyb- Thus, this acid behaves as an acid-base indicator. dochromic(II1) acids are tribasic, that 12-tungsto- The p H dependence of the 12-tungstophosphoric silicic acid is tetrabasic and that 9-tungstophos- acid is also important from the analytical point of phoric acid is hexabasic. However, the 12-molyb- view, since spectrophotometric measurements are doceric(1V) and 12-tungstophosphoricacids showed often utilized for determining the Concentration of a steady increase in basicity (number of hydrogen heteropoly compounds. These data may be m i 5 ions in solution per mole of acid) with dilution, interpreted in the case of the 12-tungstophosphoric indicating that these acids were not fully ionized acid if the pH factor is not taken into account. in the concentration range studied (-1 X to Experimental 2X M ) . According to the accepted molecular 1. Materials.--12-Tungstophosphoric acid, Q-tungstoformula, 12-molybdoceric(IV) acid, H&eMo12042, phosphoric acid, 12-tungstosilicic acid, 12-molybdoceric(1V) should be octabasic. The problem of the basicity acid and 6-molybdochromic(III) acid were prepared and purified as described earlier.44 Solutions were prepared by of the 12-tungstophosphoricacid has been discussed direct weighing and dissolution in doubly distilled water. in previous papers4J where we have shown that in 2. Basicity Measurements-Basicity determinations solution the tribasic formulation must be abandoned were made utilizing, in principle, the same method as and a higher basic formulation such as hepta- described in detail in our previous paper.4 Essentially, i t is based on measuring the optical density of an indicator in basic acid, H~PWEO~Z, must be utilized. Since X- the presence of various concentrations of the heteropoly ray analysis has indicated the P W 1 2 0 * ~ion ~ *in the acid. If the concentrations of the heteropoly acid are crystal, we assume that H7PW120a and the ionic chosen to give a pH within the critical range for a specific species obtained by its dissociation are formed in indicator, the optical density varies in a very sensitive way with the hydrogen ion activity. In our previous work we solution by hydrolysis of the tribasic species. The have used methyl orange for which we obtained a calibration question then arises as to whether additional curve over a convenient range of HC1 for the dependence of where R = [A-]/[HA], waters may be added giving anions containing 43 the indicator ratio, R, upon [E+]? oxygens, etc., in which case the experimentally ob- [A-] and [HA] being the equilibrium concentrations of the basic and the acid forms of the indicator. R can be calserved basicity might be expected to increase culated from optical densities if the optical density of the beyond 7 upon continued dilution. The same acid and the basic forms of the indicator are argument could be applied to the 12-molybdoceric- I n addition we have explored the effect of the ionic strength upon the indicator equilibrium and found that below p = (IV) acid. M the presence of neutral electrolytes does not affect I n order t o work a t lower concentrations, indi- 0.001 the equilibrium. Utilizing these data we were able to detercators having a p K higher than that of methyl mine basicities of heteropoly acids by comparing R for a orange had to be used. Accordingly, we have given concentration of a heteropoly acid with the calibration determined the basicities in the lower concentra- curve. I n the present, work, we were interested primarily in obtion range of the 12-molybdoceric(IV) and 12- taining the basicities at very low concentrations of heteropoly tungstophosphoric acids using brom cresol green, acids, hence experiments with brom cresol green (pH interval methyl red and brom cresol purple as indicators. 3.8-5.4), methyl red (4.4-6.2), and brom cresol purpie (5.2-6.8) were performed. In the concentration range of We have also determined the basicity of 9-tungsto- heteropoly acids used in these experiments (3 X I O 4 to (1) K. F. Sohulz, E. MatijeviC and M. Kerker, J . Chem. Eng. Data, in press. (2) Supported by the U. 8. Atomic Energy Commission Contract No. AT(30-1)-1801. (3) Participant in the N.S.F. Summer Research Project, 1959. (4) E. Matijevib and M. Kerker, J . Am. Chem. doc., 81, 5560 (1959). ( 5 ) E. MatijeviC and M. Kerker, TRISJOURNAL, 6!2, 1271 (1958).
1 X 10-4 M ) , the ionic ntrength effect is negligible. We have also compared the effects of various inorganic electro-
(6) E. MatijeviC and M. Kerker, J . Am. Chem. SOC.,81, 1307 (1959). (7) I. M. Klota, Thesis, The University of Chicago, 1940; see also R A. Robinson and R. H. Stokes, “Electrolyte Solutions,” Butterworth Soi. Publ., London, 1955.
.
FURTHER STUDIESON BASICITY OF HETEROPOLY ACIDS
Jan.. 1961
-
lytes and salts of heteropoly acids upon the absorption of methyl orange. Below p 0.002 they all behaved identically, indicating that there waa no specific interaction between the indicator and the heteropoly anion in the working range. Since the results obtained with indicators used in this paper overlap nicely with those found with methyl orange no specific interaction with the former was assumed. However, a number of experimental difficulties had to be overcome in order to eliminate the influence of C02 on the ionization of the indicator. We modified the experimental procedure because the reproducibility of the results obtained using a calibration curve wa5 quite poor a t the lowest concentrations. In the experiments reported here, we always compared the optical density of a solution containing a certain amount of a heteropoly acid and a constant amount of the indicator with that of a number of solutions containing exactly the same amount of indicator but varying amounts of HC1. The concentrations of HC1 were chosen 80 that the solutions had optical densities very close to that of the heteropoly acid solution. When the optical density of solutione containing the hetcropoly acid and HC! were exactly the same, the basicity of the heteropoly acid was obtained by direct comparison of molarities of both solutions. When the optical density was slightly different the basicity was calculated from the corresponding R values. Optical density measurements were determined using the Beckman Model DU spectrophotometer which was e uipped with a thermostat jacket in the cell compartment %rough which was circulated water from a 25' constant temperature bath. Calibrated 1 cm. Corex cells were used in all experiments. The working wave lengths for the different indicators were: brom cresol green 450 and 620 mp, methyl red 520 mp, and brom cresol purple 430 and 590 mp. These wave lengths gave the largest differences in optical absorption between the acid and the basic forms of the indicator. All solutions in one experiment were prepared using the same batch of clistmed water. Before taking readings, dry helium or nitrogen was bubbled through solutions for 10 minutes. This bubbling time was found sufficient to eliminate the effects of dissolved COI and to obtain reproducible readings of optical densities. Each measurement was repeated several times using at least two different cells and only results with good internal agreement were used. When two wave lengths were utilized the same procedure was repeated with each. All solutions were prepared by successive dilution of conconcentrated stock solution uang the water from the same batch. The concentration of all indicators was the same (1.876 X 10- M ) and kept constant throughout the experiments. The concentration was chosen to give optimal optical density readings. The effect of the ionization of indicator is compensated in the technique employed, since both the heteropoly acid and hydrochloric acid solutions contained the same amount of indicator a t very nearly equal activities of hydrogen ions. All pH measurements have been done using glass electrodes and Model G Beckman pH meter.
57 25'
E
s
-t
3-s m a m
4
? -4 -5 -G LO . MOLAR CONC. OF H E ~ E R O P O L YACID. Fig. 1.-Basicities of 6-molybdochromic( 111) acid, 12tungstosilicic acid, 9-tungsto hosphoric acid, 12-tungs& phosphoric acid and l%mof)ybdoceric(IV) acid against concentration of investigated solution. InJk+iE! employed: methyl oran e (MO), methyl red (MR), brom cresol green (BCG) andqbrom cresol purple (BCP).
Reisults and Discussion Fig. 1, the basicities of 6molybdochromic(II1) acid, 12-tungstosilicic acid, 9-tungstophosphoric acid, 12-tungstophosphoric acid and 12-molybdoceric acid are plotted against the molar concentration of the heteropoly acid. These data represent a combination of our earlier results4 obtained using methyl orange a t concentrations of heteropoly acids higher than 2 x 10-6 M and the new data using brom cresol green, methyl red and brom cresol purple for acids in concento 1 X M . The 6tration range: 3 X molybdochromic acid, 12-tungstosilicic acid and 9tungstophosphoric acid maintained the same basici201 I 1 1 I I I ties of 3, 4 and 6 , respectively, a t the lower con300 320 340 360 m p centrations indicating that all three acids were fully W A V E LENGTH, A . ionized in the higher concentration range and that Fig. 2.-Molar extinction coefficients of 12tungstophosno decompositioii or hydrolysis takes place upon phoric acid in its acid and basic form for the wave length dilution. range from 300 to 365 mu. 1. Basicity.--.In
58
BILLYR.
I n the higher concentration range previously investigated with methyl orange, the 12-tungstophosphoric acid and the 12-molybdoceric acid showed a definite trend toward increasing basicity with dilution. However, the basicities corresponding to structural formulas H1PW12042and H~CeM012042were not reached. By extending the measurements to lower concentrations of these acids with the aid of the indicators of higher pK, the basicity values now show further increase until they eventually level off at 7 for the 12-tungstophosphoric acid and 8 for the molybdoceric(1V) acid. Thus, it appears that both of these acids can be obtained fully ionized as free hepta- and octavalent heteropoly ions in these very dilute solutions. I n addition, these results further confirm the above formulations for these acids in aqueous solution. The data for 12-tungstophosphoric and 1Bmolybdoceric(IV) acids at the lowest concentrations suggest that the last stage of the ionization occurs stepwise. I n Fig. 1, we have delineated this by the dashed line, the full line being drawn smoothly through all the data. 2. 12-Tungstophosphoric Acid as a n Acid-Base Indicator.-In the preparation of the sodium salt of 12-tungstophosphoric acid by ion exchange, we found that although the pH increase upon passage through the column was comparable to that obtained for the 9-tungstophosphoric and 12-tungstosilicic acids, the optical density in the ultraviolet of the steady-state effluent was only a small fraction of that of the free acid. Similar effects were obtained when we attempted to prepare the am-
LOY
Vol. 65
monium and silver salts of the free acid by the same method. However, we found that if dilute nitric acid was added so that the pH of the free 12tungstophosphoric acid was restored, the optical density returned to a value nearly equal to that for the free acid. This indicates that the ionized form of the acid absorbs differently from the unionized form so that 12-tungstophosphoric acid behaves as an acid-base indicator in the ultraviolet. We obtained the molar extinction coefficient of the acid form of the 12-tungstophosphoric acid from the limit of the optical density of the acid in the presence of increasing concentration of nitric acid. The limiting value was determined from a plot of optical density against the reciprocal of the nitric acid concentration, the optical density of the nitric acid itself being taken into consideration at the highest concentrations. The results are given in Fig. 2. At very high dilutions, solutions of the sodium salt followed Beer's law, indicating that the anion was completely dissociated so that in this case the optical density was used directly to obtain the absorption index of the basic form of the indicator. These results are also given in Fig. 2. If 12-tungstophosphoric acid were monobasic, the optical density could now be used to calculate the degree of ionization. However, since 12-tungstophosphoric acid is polybasic this problem is complicated by the existence of the various intermediate species between the fully associated and fully ionized acid.
THE ACTION OF OXYGEN ON IRRADIATED POLYVINYL CHLORlDE BY BILLYR. LOY Physical Research Laboratory, The Dow Chemical Company, Midland, Michigan Received M a y 14, 1960
The formation and decay of the peroxy radical produced in irradiated poly-(vinyl chloride) is studied by electron spin resonance (e.s.r.) spectroscopy over a temperature range from -78 to +45". The two reactions are studied independently making use of the difference in reaction rates. The exponential formation reaction is thought to be diffusion controlled. Tke apparent f i r s b - o r d m h y reaction is interpreted as an autotixidation process which gives mobility to the unpaired electron.
Introduction The effect of air on irradiated poly-(vinyl chloride) has been noted by other workers. Miller' has shown by electron spin resonance (e.s.r.) measurements that there is a much more rapid loss of radicals in irradiated PVC that has been exposed to air than similarly treated samples in DUCUO. Chapiro2 and Miller1 both report a lack of coloration of the material when it is exposed to air, presumably through the formation of a peroxy radical. The e.s.r. spectrum of the peroxy radical is particularly well-suited for study (by e.s.r.). Abraham and Whiffen3 studied the formation and decay of peroxy radicals in polyethylene and polychloro(1) A. A. Miller, THIBJOURNAL, 63, 1755 (1959). A. Chapiro, J . chim. phys., 68, 895 (1956). (3) R.J. Abraham and D. H. Whiffen, Trans. Faraday SOC.,84,1298 (1958). (2)
trifluoroethylene. They suggest that decay occurs via reversible and irreversible processes. As a result of studying the formation and decay over a temperature range from -78 to +45", this article proposes that the peroxy radical sites effectively acquire motion through an autooxidation reaction. I n this work, the samples were irradiated with y-rays a t liquid nitrogen temperature. The action of oxygen on the irradiated samples was observed, in situ, by e.s.r. Experimental Material.-The polyvinyl chloride (PVC) used in these experiments was manufactured by The Dow Chemical Company and is designated PVC-111-4, Lot 11727, Blend 2. The average particle size is about 140 p , but the surface area as determined by the BET method is 1.24 m.2/g. Therefore, the particles must be regarded as extremely porous. No plasticizers, inhibitors or other additives were used in the manufacture of this material.
